Domain Generalization via Multidomain Discriminant Analysis

被引:0
|
作者
Hu, Shoubo [1 ]
Zhang, Kun [2 ]
Chen, Zhitang [3 ]
Chan, Laiwan [1 ]
机构
[1] Chinese Univ Hong Kong, Dept Comp Sci & Engn, Hong Kong, Peoples R China
[2] Carnegie Mellon Univ, Dept Philosophy, Pittsburgh, PA 15213 USA
[3] Huawei Noahs Ark Lab, Hong Kong, Peoples R China
来源
35TH UNCERTAINTY IN ARTIFICIAL INTELLIGENCE CONFERENCE (UAI 2019) | 2020年 / 115卷
基金
美国国家卫生研究院; 美国国家科学基金会;
关键词
INFERENCE; KERNEL;
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Domain generalization (DG) aims to incorporate knowledge from multiple source domains into a single model that could generalize well on unseen target domains. This problem is ubiquitous in practice since the distributions of the target data may rarely be identical to those of the source data. In this paper, we propose Multidomain Discriminant Analysis (MDA) to address DG of classification tasks in general situations. MDA learns a domain-invariant feature transformation that aims to achieve appealing properties, including a minimal divergence among domains within each class, a maximal separability among classes, and overall maximal compactness of all classes. Furthermore, we provide the bounds on excess risk and generalization error by learning theory analysis. Comprehensive experiments on synthetic and real benchmark datasets demonstrate the effectiveness of MDA.
引用
收藏
页码:292 / 302
页数:11
相关论文
共 50 条
  • [1] Generalization of discriminant analysis for possibility distributions
    Miyamoto, S
    Sato, M
    Umayahara, K
    1998 SECOND INTERNATIONAL CONFERENCE ON KNOWLEDGE-BASED INTELLIGENT ELECTRONIC SYSTEMS, KES '98, PROCEEDINGS, VOL, 3, 1998, : 177 - 182
  • [2] A GENERALIZATION OF THE DISCRIMINANT FUNCTION-ANALYSIS
    RAO, MM
    ANNALS OF MATHEMATICAL STATISTICS, 1957, 28 (04): : 1062 - 1063
  • [3] HETEROGENEOUS DOMAIN GENERALIZATION VIA DOMAIN MIXUP
    Wang, Yufei
    Li, Haoliang
    Kot, Alex C.
    2020 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING, 2020, : 3622 - 3626
  • [4] Unsupervised domain adaptation via transferred local Fisher discriminant analysis
    Mozhdeh Zandifar
    Samaneh Rezaei
    Jafar Tahmoresnezhad
    Iran Journal of Computer Science, 2023, 6 (4) : 345 - 364
  • [5] Domain Generalization via Rationale Invariance
    Chen, Liang
    Zhang, Yong
    Song, Yibing
    van den Hengel, Anton
    Liu, Lingqiao
    2023 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION, ICCV, 2023, : 1751 - 1760
  • [6] GENERALIZATION BOUNDS FOR DOMAIN ADAPTATION VIA DOMAIN TRANSFORMATIONS
    Vural, Elif
    2018 IEEE 28TH INTERNATIONAL WORKSHOP ON MACHINE LEARNING FOR SIGNAL PROCESSING (MLSP), 2018,
  • [7] Domain Generalization via Gradient Surgery
    Mansilla, Lucas
    Echeveste, Rodrigo
    Milone, Diego H.
    Ferrante, Enzo
    2021 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2021), 2021, : 6610 - 6618
  • [8] Generalization of the discriminant and applications
    El Fadil, L
    Charkani, ME
    ARABIAN JOURNAL FOR SCIENCE AND ENGINEERING, 2004, 29 (1A) : 93 - 98
  • [9] On the discriminant criterion and a generalization
    Eggar, MH
    AMERICAN MATHEMATICAL MONTHLY, 1996, 103 (09): : 788 - 792
  • [10] On the Discriminant Criterion and a Generalization
    Eggar, M. H.
    American Mathematical Monthly, 103 (09):